Design, implementation, and testing of an adaptable optimal controller for an electric wheelchair

The design, implementation, and testing of an adaptable optimal controller for an electric wheelchair are presented. Optimal control theory and pattern recognition techniques are combined to produce the design of a controller using a modified proportional, integral, and derivative (PID) control method. The controller is adaptable in the sense that multiple sets of control coefficients are used with each set being optimized for a specific range of wheelchair load parameters. The appropriate set of control coefficients may be automatically selected based on load parameter estimates to provide a self-adaptive controller, or manually selected prior to installation to tailor the chair to a particular user. The manually adaptable controller is implemented in a microprocessor-based system. The experimental results obtained from installing and testing the system on an electric wheelchair are presented and discussed.